CSF neurochemical profile and cognitive changes in Parkinson’s disease with mild cognitive impairment

www.nature.com/npjparkd ARTICLE OPEN CSF neurochemical profile and cognitive changes in Parkinson’s disease with mild cognitive impairment Federico Paolini Paoletti 1,2,3, Lorenzo Gaetani1,2,3, Giovanni Bellomo 2, Elena Chipi1, Nicola Salvadori1, Chiara Montanucci1, Andrea Mancini1, Marta Filidei1, Pasquale Nigro 1, Simone Simoni1, Nicola Tambasco 1, Massimiliano Di Filippo1 and Lucilla Parnetti 1,2 ✉ 1234567890():,; Pathophysiological substrate(s) and progression of Parkinson’s disease (PD) with mild cognitive impairment (PD-MCI) are still matter of debate. Baseline cerebrospinal fluid (CSF) neurochemical profile and cognitive changes after 2 years were investigated in a retrospective series of PD-MCI (n = 48), cognitively normal PD (PD-CN, n = 40), prodromal Alzheimer’s disease (MCI-AD, n = 25) and cognitively healthy individuals with other neurological diseases (OND, n = 44). CSF biomarkers reflecting amyloidosis (Aβ42/40 ratio, sAPPα, sAPPβ), tauopathy (p-tau), neurodegeneration (t-tau, NfL, p-NfH), synaptic damage (α-syn, neurogranin) and glial activation (sTREM2, YKL-40) were measured. The great majority (88%) of PD-MCI patients was A-/T-/N-. Among all biomarkers considered, only NfL/p-NfH ratio was significantly higher in PD-MCI vs. PD-CN (p = 0.02). After 2 years, one-third of PD-MCI patients worsened; such worsening was associated with higher baseline levels of NfL, p-tau, and sTREM2. PD-MCI is a heterogeneous entity requiring further investigations on larger, longitudinal cohorts with neuropathological verification. npj Parkinson’s Disease (2023)9:68 ; https://doi.org/10.1038/s41531-023-00509-w INTRODUCTION Cognitive dysfunctions are frequently reported in Parkinson’s disease (PD), even in newly diagnosed patients, potentially evolving into dementia1. Following the paradigm of Alzheimer’s disease (AD) clinical continuum2, the concept of mild cognitive impairment (MCI) as a prodromal stage of dementia has been also applied to PD3,4. In a recent meta-analysis, Parkinson’s disease with mild cognitive impairment (PD-MCI) has shown variable clinical presentations and progression to dementia5. Different pathophysiological mechanisms are potentially involved in PD-MCI6. Although neuropathological data devoted to this specific entity are scanty and carried out in small series, Lewy body pathology, Alzheimer pathology, and cerebral amyloid angiopathy seem to be related to PD-MCI clinical phenotypes7,8. The difficulties in identifying the neurobiological substrate(s) represent a matter of debate about the validity of the MCI construct in the clinical management of PD patients9. To give a contribution in this field, we retrospectively analyzed a cohort of patients with PD categorized as PD-MCI3 and as cognitively normal (PD-CN), for whom cerebrospinal fluid (CSF) samples collected at the diagnostic work-up were available. As a control group, we considered cognitively healthy individuals who underwent CSF analysis for other minor neurological disorders (OND). We also included, as a contrast group, patients with MCI due to AD (MCI-AD), who have, by definition, a typical CSF neurochemical profile and the highest risk of progression to dementia. In these groups we measured a large panel of CSF biomarkers reflecting different pathophysiological pathways including amyloidosis (soluble amyloid precursor protein α and β, sAPPα,β; β-amyloid 1–42/1–40 ratio, Aβ42/Aβ40), tauopathy (181-phosphorylated tau, p-tau), amyloid-dependent neurodegeneration (total-tau, t-tau), amyloid-independent neurodegeneration (neurofilament light chain, NfL; phosphorylated neurofilament heavy chain, p-NfH), synaptic damage (total α-synuclein, α-syn; neurogranin, Ng), and glial activation (soluble triggering receptor expressed on myeloid cells 2, sTREM2; chitinase-3-like protein 1, YKL-40). In PD-CN and PD-MCI patients, we also verified the longitudinal changes in cognitive functioning by considering screening tests after 2 years. The aims of our study were: (i) to investigate whether a wellcharacterized cohort of PD-MCI patients, compared to PD-CN individuals, show a differential neurochemical profile by measuring a large panel of CSF putative biomarkers reflecting different pathophysiological pathways; (ii) to verify changes in cognitive performances after 2 years, also investigating the potential predictive role of CSF biomarkers on cognitive outcome. RESULTS Demographic and clinical features We included 40 PD-CN, 48 PD-MCI, 25 MCI-AD, and 44 OND patients. The details of the main demographic and clinical features of each diagnostic group are summarized in Table 1. At baseline, the great majority of patients were drug-naïve (80% of PD-CN patients and 75% of PD-MCI patients). As expected, PD-MCI patients showed lower Montreal Cognitive Assessment (MoCA) and Mini-Mental State Examination (MMSE) scores, and higher Unified Parkinson’s Disease Rating Scale-part III (UPDRS-III) and Hoehn and Yahr (H&Y) scores, as compared to PD-CN individuals. Details referring to the baseline neuropsychological evaluation in the two diagnostic groups are reported in Table 2. At the comprehensive clinical evaluation carried out after two years, the diagnosis of PD was confirmed in all patients. Out of 88 PD patients, neuropsychological assessment at follow-up was available for 65 (28 PD-CN and 37 PD-MCI). Cognitive worsening (loss of at least 2 points at both MMSE and MoCA) was observed in 30% of PD-MCI patients, as opposed to 43% of PD-MCI patients with unchanged scores. None of PD-CN individuals worsened both 1 Section of Neurology, Department of Medicine and Surgery, University of Perugia, Perugia, Italy. 2Laboratory of Clinical Neurochemistry, Section of Neurology, Department of Medicine and Surgery, University of Perugia, Perugia, Italy. 3These authors contributed equally: Federico Paolini Paoletti, Lorenzo Gaetani. ✉email: Published in partnership with the Parkinson’s Foundation F. Paolini Paoletti et al. 2 Table 1. Demographic and clinical features in each diagnostic group. PD-CN PD-MCI MCI-AD OND PD-MCI vs. PD-CN Baseline n 40 48 25 44 – Sex (F/M) 16/24 22/26 14/11 23/21 – Age (y) 64 ± 7 68 ± 6 72 ± 5 64 ± 14 – Disease duration (y) 1.7 ± 1.1 1.9 ± 1.5 1.7 ± 0.6 – – H&Y 1.9 ± 0.7 2.1 ± 0.4 – – 0.0036 UPDRS-III baseline 23.8 ± 10.8 29.2 ± 8.2 – – 0.0048 MMSE MoCA 28.3 ± 1.5 25.1 ± 3.1 26.0 ± 2.4 18.9 ± 3.9 21.9 ± 3.3 – 28.8 ± 0.8 – <0.0001 <0.0001 Patients on dopaminergic treatment 8/40 12/48 – – – LEDD 74 ± 173 107 ± 196 – – – 35 1234567890():,; 2-year follow-up n 28 37 25 UPDRS-III score change −6.9 ± 16.6 −2.8 ± 8.2 – – – MMSE score change −0.1 ± 3.0 −1.0 ± 3.1 −3.1 ± 3.4 −0.2 ± 1.0 – MoCA score change 0.2 ± 4.3 −1.0 ± 4.2 – – 0.035 Data are expressed as mean ± standard deviation. Significance level set to 0.005 to account for multiple testing effects in PD-CN vs. PD-MCI comparison. Nonsignificant p-values are not shown. H&Y Hoehn and Yahr (...truncated)